# Cyclic Ablation: Testing Concept Localization against Functional Regeneration in AI

arXiv: https://arxiv.org/abs/2509.25220 · HTML: https://mlsanity.com/paper/2509.25220/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Probing deception, bias and beliefs, one of 13 directions in Mechanistic interpretability.
- Builds on 6 of its 9 references that are in the map.
- First posted on arXiv 23 Sep 2025.

## Placed in

- Mechanistic interpretability / [Probing deception, bias and beliefs](https://mlsanity.com/mechinterp/probing-deception-bias-and-beliefs/)

## Builds on (most cited first, 6 of 6)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022)
- [Sparse Autoencoders Find Highly Interpretable Features in Language Models](https://mlsanity.com/paper/2309.08600/) (2023)
- [Mass-Editing Memory in a Transformer](https://mlsanity.com/paper/2210.07229/) (2022)
- [The Hydra Effect: Emergent Self-repair in Language Model Computations](https://mlsanity.com/paper/2307.15771/) (2023)

## Abstract (arXiv)

Safety and controllability are critical for large language models. A central question is whether undesirable behaviors like deception are localized functions that can be removed, or if they are deeply intertwined with a model's core cognitive abilities. We introduce "cyclic ablation," an iterative method to test this. By combining sparse autoencoders, targeted ablation, and adversarial training on DistilGPT-2, we attempted to eliminate the concept of deception. We found that, contrary to the localization hypothesis, deception was highly resilient. The model consistently recovered its deceptive behavior after each ablation cycle via adversarial training, a process we term functional regeneration. Crucially, every attempt at this "neurosurgery" caused a gradual but measurable decay in general linguistic performance, reflected by a consistent rise in perplexity. These findings are consistent with the view that complex concepts are distributed and entangled, underscoring the limitations of direct model editing through mechanistic interpretability.